An assistant that only talks is safe in a boring way. The worst it can do is be wrong, and you find out immediately because you are reading the answer.
The moment it can act — send the message, make the booking, move the file, pay the invoice — the question changes entirely. It is no longer what does it know. It is what is it allowed to do, and how would I find out if it did something I did not want.
Most of this category is currently answering the first question loudly and the second one not at all.
Where we stand, in one line: It reads what you authorise and drafts on its own. It sends nothing, posts nothing and buys nothing without your yes. It never moves money and never signs anything. The whole table, row by row →
Three kinds of action, and they are not alike
Lumping them together is the root of most bad design.
Reversible and private. Reading a document, drafting a reply, organising a folder, preparing a summary. If it gets these wrong you delete the output. Nobody outside knows it happened. These can happen freely.
Irreversible but private. Deleting something, overwriting a file, cancelling a standing arrangement. Nobody else sees it, but you cannot undo it. These need a confirmation, every time, however tedious.
Anything other people see. Sending an email, submitting a form, making a payment, posting anything anywhere. Once it has gone, it has gone, and the cost is borne in front of other people — a recipient, a counterparty, a colleague acting on what was written. These need you, individually, each time — and a standing permission for this category is a mistake dressed up as convenience.
The useful rule is not about how confident the system is. It is about whether the action can be taken back and who sees it.
What a yes is worth
There is a second question underneath the first, and it is the one people miss: not whether the assistant asks, but what your answer is worth afterwards. A yes that lives forever is not consent; it is a setting you agreed to once and forgot.
The arrangement we use has three parts, and each is checkable rather than decorative:
- Consequential actions wait for you. Nothing is sent or booked without your approval — one action at a time, named plainly enough to judge.
- Approvals expire. A yes covers what it named and then lapses. There is no standing yes quietly widening in the background.
- Everything is logged. Not a feed of activity — a written record of what was done, when, and on whose approval, that you can read afterwards.
And the same discipline runs underneath the answers themselves: they cite your own records, and the provenance — source, date, location — is kept end to end. When the assistant proposes to do something, you can see the ground it stands on before you say yes.
Why “it asked me first” is not enough
Consent stops being meaningful when it is requested too often. An assistant that asks permission forty times a day trains you to approve without reading, which is worse than not asking, because now there is a record of you agreeing.
Good design therefore asks rarely and specifically. Not “allow the assistant to manage your email?” but “send this reply, to these two people, saying this?”. The first is unanswerable. The second you can judge in a second.
What to insist on
If you are evaluating anything that can act on your behalf:
A written record of every action taken. Not a feed of activity — a record you can read afterwards, showing what was done and on what basis.
An approval before anything consequential. If the honest answer to “what can it send or book without asking” is anything other than nothing, you are being asked to trust a policy rather than a design.
Approvals that expire. A yes given in March should not still be working in August. If permission never lapses, the asking was a ceremony.
Answers that show their sources. If the assistant cannot say which of your records an answer came from, you cannot judge its proposals either — and every approval you give is a guess.
The honest tension
Every constraint above costs convenience. An assistant that asked nothing and did everything would be more useful on a good day and considerably worse on a bad one, and you cannot know in advance which kind of day it is.
Our position is that the boring arrangement is correct: read and draft freely, confirm anything irreversible, and require a person for anything other people will see. It makes for a less impressive demonstration and a system you can still trust in month eight.
What to do when it gets something wrong →
Every promise, in plain words →
Your own AI. Your own material. One assistant, yours alone.
$9,995 the first year — everything included. $5,995 each year after. The first year costs more because it contains the preparation — your archive prepared and loaded by hand under a written scope, and your assistant built on it — as well as the first year of the service; every year after is the service running.
Or add the Archive Assessment first — $1,950, credited in full
You order and pay at checkout; we write back within days — a decline returns every dollar, and until our letter confirms the year you may withdraw in writing. From that letter the year is final, and it arrives on the date the letter names. Prefer to write to us first?